Polygon detection and Localization based on link-line model

Polygon detection and Localization technique could be applied in the industrial environment as many industrial objects contain some planes with polygons inside. The objects could be located when the polygons inside them have been located. In this paper, we propose a new method for polygon detection and localization. First, the Canny operator is applied to acquire the edges of the image and the line segments are extracted from the edges; Then the connection relationship of these line segments is analyzed and the so call “link-line” model is generated and the polygons are found by traversing the link-line model; At last the 6 degree-of-freedom pose (positions and orientations) of the polygons are calculated according to the equations of perspective projection and the geometry features of the polygons. The proposed method was tested on 50 images and the results demonstrated that the proposed method is effective and robust for polygon detection and localization.